Seth Darren’s Post

📚Unpacking the Naive RAG Approach🤖 The naive RAG approach depicted offers a promising avenue for question-answering systems, leveraging the power of language models with the precision of information retrieval. By vectorizing document chunks and utilizing a Vector Database, it effectively narrows down the context needed for generating accurate responses, showcasing the potential of such systems in understanding and utilizing large volumes of information. However, it's important to appreciate that this model is in its nascent stage, optimized for proof of concept rather than production environments. As such, it should be approached with caution for production use, where scalability, robustness, and error handling are critical, and further refinement and testing are necessary to meet those demands. #MachineLearning #QuestionAnswering #AILanguageModels #InformationRetrieval #NLP #ArtificialIntelligence #DataScience #POCvsProduction #InnovationInAI #VectorSearch

Call for Papers - Upcoming Issue!! International Journal on Natural Language Computing (IJNLC) ISSN : 2278 - 1307 [Online]; 2319 - 4111 [Print] Webpage URL: https://lnkd.in/gCsAdFDb Submission URL: https://lnkd.in/gqf-SCmH Here's where you can reach us : ijnlcjournal@yahoo.com or ijnlcjournal@airccse.org or ijnlc@aircconline.com #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval #bigdata #bioinformatics #datamining #machinelearning #fuzzylogic #textmining #neuralnetworks #engineering #technology #computervision #deeplearning #ontologies #socialmedia #web #analytics #ubiquitouscomputing #naturallanguageprocessing #ai #knowledgerepresentation #multimedia

Call for Papers - February Issue!! International Journal on Natural Language Computing (IJNLC) ISSN : 2278 - 1307 [Online]; 2319 - 4111 [Print] Webpage URL: https://lnkd.in/gCsAdFDb Submission URL: https://lnkd.in/gqf-SCmH Here's where you can reach us : ijnlcjournal@yahoo.com or ijnlcjournal@airccse.org or ijnlc@aircconline.com #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval #bigdata #bioinformatics #datamining #machinelearning #fuzzylogic #textmining #neuralnetworks #engineering #technology #computervision #deeplearning #ontologies #socialmedia #web #analytics #ubiquitouscomputing #naturallanguageprocessing #ai #knowledgerepresentation #multimedia

Do you want to see how to correct a 3D human pose with natural language? In this work, we introduce the PoseFix dataset, which consists of thousands 3D human pose pairs and the corresponding texts describing how to modify the source pose to obtain the target pose. We demonstrate the potential of this dataset on two tasks: (1) text-based pose editing, that aims at generating corrected 3D body poses given a query pose and a text modifier; and (2) correctional text generation, where instructions are generated based on the differences between two body poses. If you are interested, join us on Thursday (October 5th) afternoon at "Foyer Sud" - poster 097 at #ICCV2023. PoseScript: 3D Human Poses from Natural Language, Ginger Delmas, Philippe Weinzaepfel, Francesc Moreno Noguer and Grégory Rogez, ICCV 2023 Paper: https://lnkd.in/ere6_GwH dataset and code: https://lnkd.in/eA7imBEG #computervision #humanbody #naturallanguageprocessing #nlp #artificialintelligence  #NaverLabsEurope

Current Issue!! Volume URL: https://lnkd.in/gp4d6HNW June 2024, Volume 13, Number 3 Fine-Tuning of Small/Medium LLMs for Business QA on Structured Data Rasha Ragab and Abdulrahman Altahhan, University of Leeds, UK Full Text: https://lnkd.in/g5kVZkYA Human Intention Space - Natural Language Phrase Driven Approach to Place Social Computing Interaction in a Designed Space Pronab Pal, KeyByte Systems, Australia Full Text: https://lnkd.in/gYJ2_ke3 #phonology #spokenlanguage #naturallanguageprocessing #ai #knowledgerepresentation #multimedia #nlpresearchpapers #nlp2024 #nlparticles #currentissuenlp #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval

Call for Papers - June Issue!! ~~~~~~~~~~~~~~~~~~~~~~ International Journal on Natural Language Computing (IJNLC) ISSN : 2278 - 1307 [Online]; 2319 - 4111 [Print] Webpage URL: https://lnkd.in/eBUukCPu Submission URL: https://lnkd.in/gqf-SCmH Here's where you can reach us : ijnlcjournal@yahoo.com or ijnlcjournal@airccse.org or ijnlc@aircconline.com Google Scholar: https://lnkd.in/gsM8qyJA Wikicfp URL: https://lnkd.in/g-KfcrZA #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval #bigdata #bioinformatics  #datamining #machinelearning #fuzzylogic #textmining #neuralnetworks #engineering #technology #computervision #deeplearning #ontologies #socialmedia #web #analytics  #ubiquitouscomputing #naturallanguageprocessing #ai #knowledgerepresentation #multimedia #nlpresearchpapers #nlp2024 #nlparticles #currentissuenlp

Call for Papers - June Issue!! ~~~~~~~~~~~~~~~~~~~~~~ International Journal on Natural Language Computing (IJNLC) ISSN : 2278 - 1307 [Online]; 2319 - 4111 [Print] Webpage URL: https://lnkd.in/eBUukCPu Submission URL: https://lnkd.in/gqf-SCmH Here's where you can reach us : ijnlcjournal@yahoo.com or ijnlcjournal@airccse.org or ijnlc@aircconline.com Google Scholar: https://lnkd.in/gsM8qyJA Wikicfp URL: https://lnkd.in/g-KfcrZA #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval #bigdata #bioinformatics #datamining #machinelearning #fuzzylogic #textmining #neuralnetworks #engineering #technology #computervision #deeplearning #ontologies #socialmedia #web #analytics #ubiquitouscomputing #naturallanguageprocessing #ai #knowledgerepresentation #multimedia #nlpresearchpapers #nlp2024 #nlparticles #currentissuenlp

Call for Papers - June Issue!! WikiCFP URL: https://lnkd.in/g-KfcrZA ~~~~~~~~~~~~~~~~~~~~~~ International Journal on Natural Language Computing (IJNLC) ISSN : 2278 - 1307 [Online]; 2319 - 4111 [Print] Webpage URL: https://lnkd.in/eBUukCPu Submission URL: https://lnkd.in/gqf-SCmH Here's where you can reach us : ijnlcjournal@yahoo.com or ijnlcjournal@airccse.org or ijnlc@aircconline.com Google Scholar: https://lnkd.in/gsM8qyJA Youtube URL: https://lnkd.in/eSE_h5zn #nlp #textmining #datamining #machinetranslation #languageprocessing #machinelearning #sentimentanalysis #artificialintelligence #informationretrieval #bigdata #bioinformatics  #datamining #machinelearning #fuzzylogic #textmining #neuralnetworks #engineering #technology #computervision #deeplearning #ontologies #socialmedia #web #analytics  #ubiquitouscomputing #naturallanguageprocessing #ai #knowledgerepresentation #multimedia #nlpresearchpapers #nlp2024 #nlparticles #currentissuenlp

When building a language translation model, I was using a simple model I decided to compare it with a SOTA model from Meta But my basic model performed 5x better than the SOTA model Here’s why → My dataset focused on minority languages with lesser data, unlike the benchmarks used for major languages in Meta’s model. These languages have 10x more data than the dataset I was using. By focusing on your specific use case, even simple algorithms can outperform fancy models #nlp #machinelearning #artificialintelligence